From LTCM to Amaranth, Lessons We Learned

**lzhang** Mon Dec 11, 2006 7:53 pm

From LTCM to Amaranth, Lessons We Learned
It was this video of an MBA talk Mr. Buffett gave at the University of Florida on 10/15/1998 that reminded me the stories of Long-Term Capital Management (LTCM) and Amaranth Advisors. Both stories are well-known and have similarities. As value investors, it is good if we can learn lessons from our mistakes, but it will be even better if we could learn lessons from other people's mistakes.

LTCM was founded in 1994 with more than $1B capital raised from investors. In its team, it had 16 people with very high IQ, total of 300 or 400 years of experience in finance, including two Nobel Price winners. They had a fantastic annual return of 40% from 1994 until they went broke in late Summer/Fall 1998 with $4.6B loss, which made its total return for those years negative.

Amaranth Advisors was founded in 2000. Though it called itself multi-strategy hedge fund, it was heavily involved in natural gas trading. Eight years later after LTCM collapsed, about the same time of the year 2006, Amaranth went broke with a total loss about $6.5B.

People at LTCM and Amaranth Advisors were highly intelligent, but why did they all go broke like fools? As Mr. Buffett said, they were smart people doing dumb things. Those people were pulling triggers against their temples and hoping they were lucky so that they could miss the 1 bullet out of 1 million chambers and make a fortune. They made it a few times, but they finally hit the bullet, though it might be a six-sigma event, and they were done.

The businesses they were doing were just fine, trying to bet on things that should be normal to happen in the financial area. For example, LTCM bet on that bond prices of 'off-the-run' and 'on-the-run' tended to converge. Amaranth Advisors bet on that the spreads of natural gas contracts for Summer and Winter would widen. Those are normal under normal business situation and sound like sure bets. The problem is that since those situations are expected, even there is profit to make through those bets, they are so small that the total return will be totally minute. Now comes the solution: using leverage. With leverage up to 30:1, LTCM could transfer 1% return into 30%. In Amaranth, they used up to 8:1 leverage to bet on natural gas future price.

If it was not because of the panic caused by Russian Bond default in 1998, LTCM were confident they were not afraid of six-sigma event. If it was not because of the mild weather in Winter 2005 and summer 2006, Amaranth Advisors' strategy might just work fine assuming bull or neutral natural gas market. With this kind of excessive leverage, the once safe bets become very dangerous ones.

The lesson here is obvious: do NOT use large leverage. I'd like to further say, do not use any leverage at all, unless for very short-term transactional purpose. For (individual) investors, as Mr. Buffett claimed, it does not make any difference. But the extra risk people are exposed to due to leverage, even moderate one, is quite large and unpredictable. In unfavorable situation, no matter how remote that could be, the portfolio would be totally wiped out. The once sound value investment is not sound any longer, if the leveraged situation is not evaluated properly, which is quite hard to do, because you can never rule out the wipe out situation due to leverage. Personally, Mr. Graham went broke in the Great Depression due to his leverage level. His lesson was well learned so that we have two masterpieces in investment, Security Analysis and Intelligent Investor.

Lesson two looks like: we need margin of safety. Six-sigma event sounds like very remote thing to happen, but the catch here for LTCM is that: 1) high leverage makes the damage of very small probability event very large; 2) six-sigma event probability is from their own calculation based on their own model, which might be faulty, so the actual probability for such worst case scenario to happen might be considerably high; 3) for whatever reason it happens, it happens, and LTCM is done with such leverage no matter what, even six-sigma will not save them. "The market can stay irrational longer than you can stay solvent", very true, especially if highly leveraged.

Ref links:

http://webcompilation.googlepages.com/webcompilation.htm

Warren Buffett MBA Talk at University of Florida (October 15, 1998)

http://en.wikipedia.org/wiki/Long-Term_Capital_Management

http://www.sjsu.edu/faculty/watkins/ltcm.htm

http://en.wikipedia.org/wiki/Amaranth_Advisors

http://bloomberg.com/apps/news?pid=20601109&sid=aqLaijlpyfEA

Original post @ http://hepinvestor.blogspot.com/2006/12/from-ltcm-to-amaranth-lessons-we.html